Scale Economies and the Geographic Concentration of Industry
In recent empirical literature on spatial agglomeration, many papers find evidence consistent with location-specific externalities of some sort. Our willingness to accept evidence of agglomeration economies depends on how well key estimation problems have been addressed. Three issues are particularly troublesome for identifying agglomeration effects: unobserved regional characteristics, simultaneity in regional data, and multiple sources of externalities. Two empirical results appear to be robust to problems created by the first two issues: (a) individual wages are increasing in the presence of more-educated workers in the local labor force, which is consistent with localized human-capital externalities, and (b) long-run industry growth is higher in locations with a wider range of industrial activities, which suggests that firms benefit from being in more diverse urban environments. Other evidence is supportive of agglomeration effects related to regional demand linkages and short-run, industry-specific externalities.
Hanson, Gordon H. "Scale Economies and the Geographic Concentration of Industry." Journal of Economic Geography 1, 3 (July 2001): 255-76.